Hi, for what it’s worth that looks like an error or anomaly in the programme. With power or data issues I would have expected to see a more random effect.
How often are you updating the pixels?
What is generating the colour, the ESPs?
What is requesting the colour, some form of external influence?
Are colours requested on each ESP independently or can each be set differently?
What is special about 30 seconds?
Do any of the loops if divided by common numbers (128, 256 or 65,535) come to 30 seconds?
You mention that you have swapped some components, where the results the same?
angus40 wrote:Have you tried reducing the length/# of pixels to see what you can control stable /successfully ?
How about measuring the output DC voltage from the power supply ?
Yes. It has been a while since I did the test, but I think I was able to control 50 of them without it going erratic.
Measuring the DC voltage was showing stable at around 12v, both when measuring the posts themselves, and measuring at the end of the 25' wire run.
Barnabybear wrote:Hi, for what it’s worth that looks like an error or anomaly in the programme. With power or data issues I would have expected to see a more random effect.
How often are you updating the pixels?
What is generating the colour, the ESPs?
What is requesting the colour, some form of external influence?
Are colours requested on each ESP independently or can each be set differently?
What is special about 30 seconds?
Do any of the loops if divided by common numbers (128, 256 or 65,535) come to 30 seconds?
You mention that you have swapped some components, where the results the same?
Regarding program anomaly, yes this is what it looked like to me at first too, but switching to a shorter length of wire removes the problem. The effect is quite random, various groups of LEDs change colors regularly, despite the fact that the program should be holding steady at a solid color.
Updating Pixels. Not sure what the refresh rate is, but the color should be held solid so there should not be any change in data regardless of frequency.
Generating the color. I'm using D1 minis loaded with the WLED library on it. https://github.com/Aircoookie/WLED
Requesting the color. I control the D1 minis using the WLED android app over wi-fi.
Independent Request. Yes, they are currently set for independent control, although they can be combined at the software level if I wish. They can also be added to an xLights universe, although I haven't got that far yet.
Special about 30 seconds. Your guess is as good as mine. At first I assumed that was when they hit full power requirements, but the "warm up" shouldn't take that long, only about 10 seconds.
Loops Divided. I'm actually uncertain what you mean here, so I'll need more explanation to answer that question.
Swapping components. Yes, the results are always the same except when I swap it out to a shorter wire run.
Hope that helps. I appreciate any and all suggestions as I'm a bit at wits end here.
I've been doing my best to do research on stuff like this, and "ground noise" keeps coming back up. Could ground noise be causing a problem with the data wires?
Guy wrote:Yes. It has been a while since I did the test, but I think I was able to control 50 of them without it going erratic.
Measuring the DC voltage was showing stable at around 12v, both when measuring the posts themselves, and measuring at the end of the 25' wire run.
You may require power injection or if you haven't done already,disable automatic brightness limiter within the Wled gui under led preferences .
Wled is excellent and has many features . There are 2 methods to adjust brightness/intensity which may help improve your effects also . 1 directly below the color wheel and the other a slide in the top right .
Hi, thanks for the answers. It’s cool that you can’t answer all the questions; personally I think that one of life’s finest attributes is to say I don’t know, but I digress. I’m UK so it’s going to be bed time soon, mentioning this so you don’t think I’m not interested, as we all work out of the same book I’m sure someone will continue the investigation if not I’ll check in tomorrow.
Sorry more questions.
The first one I might have missed; is it the 400 or the 450 LED run that you are having the issue with?
What is the distance between the ESP D1s and the first pixel of the strings?
What brightness are you using?
Have you tried a lower brightness?
MikeKrebs wrote:Does WLED continuously update the data? Or does it send one command until it is asked to change it?Are the two data lines the exact same length? Do both carry connected ground wires?
The pattern exhibited - is it the same pattern every time it happens?
It really looks like something is generating a specific signal rather than some random noise issue.
Of course, if shorter wire cures the problem, it seems you have your answer there. Have you tried putting in a "null" pixel between the d1 and the string?
I'm not sure WLED's function regarding how often it updates data. I'll see if I can find that information.
Yes, the two data lines are same length (or within an inch at least).
The pattern is not really a pattern. It is groups of LED's switching to random colors at random times. It looks like it is a pattern but after watching it for a long enough time (I've been troubleshooting this for months on and off) I don't think it is any set pattern.
I would agree about the specific signal except for the reduction in wire length removing the problem.
One of the ideas I had in my head was to solder in a single pixel halfway through the 25' line. My thought was it might help "clean up" the noise, or at least make the ground distance shorter before it got a boost. But since I don't actually know electrical that well I thought I might just be grasping at straws. 🙂
Rainy weather and long workdays mean that I haven't been able to do any testing, but I'll get back to the other responses as soon as I can. Thanks to everyone for the help!
Guy wrote:I'm not sure WLED's function regarding how often it updates data. I'll see if I can find that information.
Yes, the two data lines are same length (or within an inch at least).
The pattern is not really a pattern. It is groups of LED's switching to random colors at random times. It looks like it is a pattern but after watching it for a long enough time (I've been troubleshooting this for months on and off) I don't think it is any set pattern.
I would agree about the specific signal except for the reduction in wire length removing the problem.
One of the ideas I had in my head was to solder in a single pixel halfway through the 25' line. My thought was it might help "clean up" the noise, or at least make the ground distance shorter before it got a boost. But since I don't actually know electrical that well I thought I might just be grasping at straws. 🙂Rainy weather and long workdays mean that I haven't been able to do any testing, but I'll get back to the other responses as soon as I can. Thanks to everyone for the help!
Your idea of adding a pixel to boost the signal is a good one, however, you would be better off using a signal amplifier (known as null pixel buffer, mu-amp or f-amp, see http://www.hansonelectronics.com.au/product/null/ for a good understanding) that boosts much more while keeping the signal clean. It also has the benefit of not requiring being nulled-out in your setup. You might even be able to put it at the beginning of your wire run and be successful. I have had similar issues but solved mine by moving the controller closer to the props.
ezellner wrote:Your idea of adding a pixel to boost the signal is a good one, however, you would be better off using a signal amplifier (known as null pixel buffer, mu-amp or f-amp, see http://www.hansonelectronics.com.au/product/null/ for a good understanding) that boosts much more while keeping the signal clean. It also has the benefit of not requiring being nulled-out in your setup. You might even be able to put it at the beginning of your wire run and be successful. I have had similar issues but solved mine by moving the controller closer to the props.
Oh! Excellent. I thought I had heard the term "null pixel" before, but couldn't really figure out where I had heard it. Glad to know I'm not crazy. Apparently it is even on the wiki here at DIY Christmas, now that I know what to search for.
I like the null pixel buffer from Hanson. Watched the video and it looks like it might do the trick. I might try it with a pixel first (since I have extras of those and therefore could do it today) and the WS2815 is a dual data line setup so I'd have to either try to splice both data lines in or get two of the controllers. Good to know that it works better if it is placed earlier in the line, I was going to split it down the middle but I might put it in at about the 10ft mark (3m) based on what Hanson is recommending.
Aside from not having to null out the first pixel in the setup, is there any other advantage to using the Null Pixel? I don't see how it could keep the line any more clean than a pixel could, but I'm still in the early learning stages of most of this stuff.
Thanks for the suggestions, I'll try to do this today and report back to everyone!
Guy wrote:Oh! Excellent. I thought I had heard the term "null pixel" before, but couldn't really figure out where I had heard it. Glad to know I'm not crazy. Apparently it is even on the wiki here at DIY Christmas, now that I know what to search for.I like the null pixel buffer from Hanson. Watched the video and it looks like it might do the trick. I might try it with a pixel first (since I have extras of those and therefore could do it today) and the WS2815 is a dual data line setup so I'd have to either try to splice both data lines in or get two of the controllers. Good to know that it works better if it is placed earlier in the line, I was going to split it down the middle but I might put it in at about the 10ft mark (3m) based on what Hanson is recommending.
Aside from not having to null out the first pixel in the setup, is there any other advantage to using the Null Pixel? I don't see how it could keep the line any more clean than a pixel could, but I'm still in the early learning stages of most of this stuff.
Thanks for the suggestions, I'll try to do this today and report back to everyone!
Note that null pixel refers to a pixel in line that is not used/illuminated. The amplifier chip is sometimes called a "null pixel without a pixel." A null pixel has to be accounted for in your setup otherwise your string will be one pixel off. The amplifiers don't have that same issue because they pass along the entire stream of data versus a pixel that strips off one pixel's worth of data and sends the rest on. The circuitry in a pixel isn't designed to send the data signal long distances whereas the amplifiers are. Many also are configured to make sure they cleanup the signal on the way through ensuring that any noise on the line gets cleaned up to a degree. That's why it's suggested they go in near the controller so any noise will be minimal.
OK, I wired a single pixel about 10' from the power/data source and 15' from the rest of the LED strip. The single pixel is behaving like I want it to, but the rest of the strip turns on random colors (mostly blue/green/white) throughout the entire strip. WLED is programmed to only light up the first 15 pixels, but the entire string turns those colors.
I am able to control the single pixel I have in the middle of the wire, but none of the others. If I disconnect the strip they all go out (obviously) and if I plug it back in they remain dark.
Still seems like a wiring issue, but the wiring simply isn't that complex!
Any other suggestions on things to test? Is there a good way to test if there is interference on my wires? I'll look to order one (two?) of those null pixels, but who knows how long they will take to get here. 🙂
-Guy
